The correction of mismatched DNA bases within recombination intermediates is thought to be a major pathway by which gene conversion occurs. In eukaryotes, gene conversion may be a key mechanism for maintaining genetic stability among gene families. Mismatch correction is also of medical interest by virtue of its ability to reduce mutation rates; the potential for activation of cellular oncogenes by single base changes underscores this contribution. However, knowledge of mismatch correction in eukaryotes is limited. The long-term goal of this project is a detailed biochemical characterization of DNA mismatch correction in the yeast S. cerevisiae. The focus will be on the pms- dependent pathway because this pathway is known to affect gene conversion, the effects are consistent with mismatch repair function, and pms mutants have been isolated. The specific aims are: 1) to characterize cis-acting factors that are likely to affect reactivity, such as single-strand breaks and different mispairs; 2) to develop active extracts of meiotic cells and test pms-dependence; 3) to isolate and characterize PMS1 protein and a protein homologous to E. coli MutS (mismatch recognition); and 4) to identify other required factors and begin their isolation. An in vitro assay for yeast mismatch correction will be based on the successful prototype used for the E. coli system. Heteroduplex DNA containing a single base mispair is prepared such that correction of the mismatch can be monitored by restriction enzyme cleavage. Covalently closed, nicked or gapped DNA will be tested for correction by yeast extracts, as will the reactivity of heteroduplexes containing different single base mispairs. The efficiency and strandedness of the reaction due to these modifications will be analyzed. Extracts will be prepared from cells induced for meiosis with the expectation that activity may be enhanced under these conditions. The dependence of this activity on pms1,pms2, and pms3 will be determined by assaying extracts from strains deficient in these gene products. PMS1 protein will be overproduced and isolated by assaying for complementation of defective extracts. A protein will be overproduced and isolated by assaying for complementation of defective extracts. A protein homologous to MutS will also be isolated. These factors will be tested for activities expected to be involved in correction such as mispair recognition or nuclease function. Other required proteins will be identified by screening extracts of strains deficient in DNA metabolism (e.g. DNA polymerization) for mismatch correction activity. The proteins involved will be isolated. A direct selection of mismatch repair mutants will also be performed.